432
J.D. Tenhunen et al.
under conditions of climate change, for example at elevated atmospheric
CO2 levels (Jarvis 1987). Essential are so-called "phenomenological" or
"aggregate models" of tundra ecosystem CO2 exchange that are simple,
based on detailed process understanding, and include appropriate "responsiveness" (Rathstetter et al. 1992).
The detailed "bottom-up" model GAS-FLUX is described as developed
for low stature tundra communities. The model integrates information on
microclimate, species composition, spatial structure, and gas exchange. A
similar model was developed by Miller et al. (1984) as a submodel driver for
a simulator of tundra ecosystem processes (ARTUS). This earlier model had
two main disadvantages that limited the potential for general application in
other ecosystems as well as at different tundra sites. First, gas exchange
physiology was described with empirical functions that are difficult to relate
to biochemistry and available nutrients, thus posing great difficulties as soon
as ecosystem-level feedbacks might be considered that modify nutrient availability. Secondly, vegetation structure was to a great extent ignored due to
low leaf area index (LAI) in tundra regions, with all vascular plant material
considered to be exposed to incident light and the total vascular plant
material functioning as a neutral filter of changing density when determining
incident light on the moss layer. The main disadvantage of this approach is
that results become questionable as soon as LAI increases above 1.0, and
the model is difficult to apply in other ecosystems of high LAI (e.g.,
shrublands or forests) without major modification.
An important advantage of the GAS-FLUX model is that estimates of gas
exchange rates are based on mechanistic descriptions of photosynthetic
processes (Harley and Tenhunen 1991) as well as on the strong correlation
found between CO2 assimilation rate and stomatal conductance. This modeling approach results in ecosystem "responsiveness" to atmospheric CO2
concentration in terms of both CO2 uptake and water loss (cf. Tenhunen
et al. 1990). A new aspect of the model is that gas exchange from the
moss component of the ecosystem is similarly treated, as moss is extremely
important in tundra regions (Tenhunen et al. 1992). The low stature of
tundra communities and limited depth to permafrost provide advantages
for model verification by chamber measurements (Gillespie et al. 1993).
The imperative need for verification of such simplified ecosystem-level
models, for development of understanding of interactions at the vegetation/
atmospheric interface, and for appropriate scaling up and adoption of common analytical philosophies among vegetation and atmospheric scientists
strongly suggests that such formulations should be constructed in an interdisciplinary setting, e.g., in the case of gas flux models, development together
with atmospheric scientists. Further verification, adjustment, and development of GAS-FLUX as a comparative analytical tool requires that structure,
leaf-level gas exchange, large chamber experiments, xylem sap flow, and
atmospheric estimates of gas exchange with the vegetation be studied simultaneously at selected sites in different climate zones. Illustrated below is the
J.D. Tenhunen et al.
under conditions of climate change, for example at elevated atmospheric
CO2 levels (Jarvis 1987). Essential are so-called "phenomenological" or
"aggregate models" of tundra ecosystem CO2 exchange that are simple,
based on detailed process understanding, and include appropriate "responsiveness" (Rathstetter et al. 1992).
The detailed "bottom-up" model GAS-FLUX is described as developed
for low stature tundra communities. The model integrates information on
microclimate, species composition, spatial structure, and gas exchange. A
similar model was developed by Miller et al. (1984) as a submodel driver for
a simulator of tundra ecosystem processes (ARTUS). This earlier model had
two main disadvantages that limited the potential for general application in
other ecosystems as well as at different tundra sites. First, gas exchange
physiology was described with empirical functions that are difficult to relate
to biochemistry and available nutrients, thus posing great difficulties as soon
as ecosystem-level feedbacks might be considered that modify nutrient availability. Secondly, vegetation structure was to a great extent ignored due to
low leaf area index (LAI) in tundra regions, with all vascular plant material
considered to be exposed to incident light and the total vascular plant
material functioning as a neutral filter of changing density when determining
incident light on the moss layer. The main disadvantage of this approach is
that results become questionable as soon as LAI increases above 1.0, and
the model is difficult to apply in other ecosystems of high LAI (e.g.,
shrublands or forests) without major modification.
An important advantage of the GAS-FLUX model is that estimates of gas
exchange rates are based on mechanistic descriptions of photosynthetic
processes (Harley and Tenhunen 1991) as well as on the strong correlation
found between CO2 assimilation rate and stomatal conductance. This modeling approach results in ecosystem "responsiveness" to atmospheric CO2
concentration in terms of both CO2 uptake and water loss (cf. Tenhunen
et al. 1990). A new aspect of the model is that gas exchange from the
moss component of the ecosystem is similarly treated, as moss is extremely
important in tundra regions (Tenhunen et al. 1992). The low stature of
tundra communities and limited depth to permafrost provide advantages
for model verification by chamber measurements (Gillespie et al. 1993).
The imperative need for verification of such simplified ecosystem-level
models, for development of understanding of interactions at the vegetation/
atmospheric interface, and for appropriate scaling up and adoption of common analytical philosophies among vegetation and atmospheric scientists
strongly suggests that such formulations should be constructed in an interdisciplinary setting, e.g., in the case of gas flux models, development together
with atmospheric scientists. Further verification, adjustment, and development of GAS-FLUX as a comparative analytical tool requires that structure,
leaf-level gas exchange, large chamber experiments, xylem sap flow, and
atmospheric estimates of gas exchange with the vegetation be studied simultaneously at selected sites in different climate zones. Illustrated below is the